5G Mobile Network Operators (MNOs) are transforming their business in two fundamental ways. First, by switching from the physical network function (PNF) approach to network deployment, to a virtualized approach (VNFs and CNFs). All of this is in service of the larger, and more strategic, second transformation: Increasing the monetization opportunities of their networks and accelerating return on their capital investments. With these changes, 5G MNOs and Network Service Providers will now be able to offer a differentiated and unique value proposition in the form of Edge Cloud services and MEC (Multi-Access Edge Computing).
In the last few years, we have seen a focus on making networks capable of supporting or hosting application workloads and their life cycles for various enterprise use cases and ISV platforms and services. The goal being to create a dedicated network slice per application or a specific service chain for an industry use case. But this has often been overshadowed by the pursuit of specific use case validation regarding the fundamental value propositions edge computing should deliver; i.e. low latency and proximity to the User Equipment (UE).
5G should be inspiring a generation of developers to create a new era of applications that can take advantage of the capabilities the MNOs have to offer. Developers have the potential to utilize the network exposure function (NEF) for the creation and utilization of service APIs composed for a given use case, and included in an application’s implementation. From its capacity to provide massively increased connectivity for IoT devices, to its super-low latency performance, and the ability to run workloads closer to where the data is being generated, the opportunities of 5G are endless.
The idea of exposing network APIs for developers to leverage is certainly in the spirit of monetizing the network asset. However, it doesn’t help the enterprise application developer actually build their applications. By definition, it’s exposing a subset of network information that an application could find useful if knowledge of the network is important, not necessarily addressing the needs of an application developer to build for the 5G edge. In fact, many approaches to edge applications have been concerned with the orchestration of workloads (containers). This is ok, but just because a workload is portable when distributed in a container, doesn’t mean it makes sense to run at the 5G edge. This approach to edge applications can't be correct in isolation, as it lacks state.
Most applications need to manage state, or act upon data in some way, so stateless workload orchestration needs to be complemented with a way for applications to act on the data. Therefore, the next step in 5G and the edge cloud service evolution has to be to provide a capability beyond workload hosting. Network service providers need to offer value in the form of a platform that enables enterprises to build edge native services that are part of their multi-cloud or hybrid cloud applications.
To build edge applications, we must not lose sight of the original idea of why the edge. The edge is born from the idea that data has gravity. If our use cases demand that we generate insights in real-time, we need to bring our insight-generating workloads (AI/ML) to where the data is. However, placing the workload is not enough. We must first address the data question and provide the mechanisms needed to intercept, capture, prepare, and process data in real-time.
But what about the fan-out nature of the network edge? How can you maintain state across a highly-distributed edge network, while also preserving the real-time goal of the edge? Well, this has to be part of the 5G Cloud. It has to replicate state within ms across the distributed points of presence (PoPs) of the network so that, as the edge application’s functions and workloads/microservices are detected, scaled, retired, or migrated (due to mobility), they can be orchestrated accordingly with their state intact.
5G MNOs need to provide their enterprise customers with a platform that is built for low latency, high-velocity data ingestion, real-time processing, insights, and actuation to inspire the creation of 5G edge cloud applications. Perhaps most of all, MNOs need to inspire a developer community with a familiar cloud experience for their application development needs.
Rather than thinking about cloud native, we should be thinking in terms of 5G edge native, what that looks like, and how to put it in the hands of developers.
